Step 1 · Try the demo
See a live semantic model at work. Nothing to install.
(opens in a new tab)
Connect
Connect your own tool to the same model
Same cubes, same measures, same numbers as the dashboard: 7,001,619 flights and a 76.3% on-time rate in 2025.
- In Excel: Data → Get Data → From Database → From Analysis Services.
- Server name:
https://demo.emondrian.com/xmla - Select the OnTime cube and click Finish.
- Build a PivotTable: drag Flights and On-Time Rate to Values and Year to Rows. 2025 shows the same numbers as the dashboard.
No Excel at hand? Open the live dashboard instead.
You should see for 2025
- flights
- 7,001,619
- on time
- 76.3%
The same numbers as the dashboard, Power BI and the agent: one model, one definition of each measure.
If a Power BI visual doesn't work as expected, see DAX support and let us know.
- In Power BI Desktop: Get data → Analysis Services.
- Server:
https://demo.emondrian.com/xmla - Choose Connect live and click OK.
- Select the OnTime cube. Add a card with Flights, filter Year to 2025: 7,001,619, the same as the dashboard.
Want to see what Power BI sends? Open Performance Analyzer in Power BI Desktop and copy the DAX query of any visual.
Power BI connects to eMondrian through the Analysis Services connector. Microsoft does not officially support this for third-party servers. See what to expect.
Want Claude to build the report for you? Step by step: Claude builds a Power BI report on this demo.
Claude (claude.ai or the desktop app)
Customize → Connectors → add a custom connector
- URL:
https://demo.emondrian.com/emondrian/mcp - Authentication: No sign-in
Then ask Claude a question about the demo data, or open Artifacts → Dashboard (Beta) and ask it to build a Claude Dashboard.
Claude Code
claude mcp add --transport http emondrian-demo https://demo.emondrian.com/emondrian/mcp
Other MCP clients
Streamable HTTP:
https://demo.emondrian.com/emondrian/mcp
Older clients that only support SSE:
https://demo.emondrian.com/emondrian/mcp/sse
On the demo the agent can list catalogs, read schemas and the Power BI model, and run MDX and DAX queries. It can also build two kinds of dashboards: Power BI reports (PBIP files with a live connection to the demo) and eMondrian Dashboards, saved on the demo with a link to open them (see Build your own dashboard). The data stays read-only: tools that change schemas or read server logs are switched off, and the data is public sample data, so it's safe to connect.
Example prompts
- “Which catalogs and cubes are on this server?”
- “In the OnTime catalog, compare the cancellation rate of American Airlines (AA) and Delta (DL) and explain the difference.”
- “Show the average arrival delay by day of week and explain the pattern.”
- “In FoodMart, what is the total Org Salary in the HR cube?”
- “Build a dashboard for the OnTime cube: on-time rate by year, by month for 2025, and the largest carriers, with year and carrier filters.”
Works with Claude Dashboards (beta) in claude.ai: Claude builds a live dashboard that queries the demo every time you open it.
Want the full tool set, including schema authoring? Run Community locally and connect to
http://localhost/emondrian/mcp (or
/emondrian/mcp/sse for SSE-only clients). In Community, the agent
signs in and acts as that user once you switch sign-in on. See the
Quickstart.
An AI agent builds a dashboard over the demo cubes and gives you a link: charts, pivots and filters that query eMondrian live in your browser. The agent needs one MCP server, eMondrian's: the same one as in the AI agent tab. It reads the cubes, checks the MDX and saves the dashboard. If you have already connected it, skip to the prompt below.
Claude Code
claude mcp add --transport http emondrian-demo https://demo.emondrian.com/emondrian/mcp
Claude (claude.ai or the desktop app)
Customize → Connectors → add a custom connector with No sign-in:
https://demo.emondrian.com/emondrian/mcp
Then ask, for example: “Build a dashboard of US flight punctuality for 2025 from the OnTime cube: flights and late arrivals by origin state, the on-time rate by month, and an airline filter.” The agent checks every query on the demo before it saves and answers with the link.
See an example first: US flight punctuality. All dashboards are listed at demo.emondrian.com/dashboard.
On the demo, dashboards are open: everyone sees every dashboard and can change or delete it, so keep to the sample data.
Server
About the demo server
What's on it
- FoodMart: the classic Mondrian retail sample (sales, inventory, and HR).
- OnTime: US flight data on ClickHouse.
- DAX module enabled, so you can try Power BI; MCP enabled for AI agents (Streamable HTTP and SSE): they read the data and build Power BI reports and eMondrian Dashboards.
- eMondrian Dashboard: dashboards an AI agent builds over these cubes through MCP.
Keep in mind
- It's a shared public server: performance is not representative.
- Data is read-only and reset regularly.
- Don't send confidential data or queries.
For developers: query the cubes directly
-
OLAP Client:
a pivot table and an MDX editor. Connect to
https://demo.emondrian.com/xmlawith no user name or password, then pick a cube. - Schema Editor: see how the cubes, dimensions and measures are defined. Read-only on the demo.
Seen enough? Try it on your data.
Install Community with Docker, connect your database, and get a free evaluation key for DAX, synthetic catalogs and dashboards.